JP4431949B2 - Red-eye correction method and apparatus for carrying out this method - Google Patents

Description

  The present invention relates to a red-eye correction method that corrects red-eye generated in captured image data, a red-eye correction program, and an image processing apparatus that implements this method.

  When taking a stroboscopic shot of a person or animal as a subject, a part of the stroboscopic light is reflected by the blood vessels in the eyeball and returned to the camera side, so the center of the pupil in the shot image is red or a different color from the actual The so-called red-eye phenomenon may occur. With the spread of digital cameras and the spread of film scanners that digitize photographic film images, many proposals have been made to solve this red-eye problem with image processing techniques. Of course, this correction work can also be done by craftsmanship technology that corrects the color tone of the pixel to the actual color while visually checking the pixel of the red eye part through the monitor, but the correction work is complicated and skilled However, it is not common considering the need for For this reason, automated techniques such as extracting the pupil position using pattern recognition technology, etc., or extracting the red part in the image and color-converting this part are also considered, but the red-eye area is accurately detected If it tries to do so, the image processing technique becomes extremely complicated, and the apparatus itself becomes expensive.

  For example, a region including an eye region that has a poor color tone is first specified, and in that region, using a feature amount that combines lightness and redness, a valley is formed between the red eye portion and the adjacent white eye and skin portion. There is a technique for separating the red-eye part from the white-eye part and the skin part by dividing the area for each mountain of the feature amount by using what can be done (for example, see Patent Document 1). In this technology, since the red eye part has a strong reflection from the retina at the center of the pupil part, the brightness tends to decrease from the center toward the peripheral part, so that the brightness including the catch light is distributed in a mountain shape, In addition, it is utilized that the iris part is a valley with the pupil part that has red eyes with respect to the magnitude of the redness value in the blue-eye pupil. That is, it is not a technique for distinguishing red-eye pixels by detecting red pixels, white pixels, or skin-colored pixels, trying to distinguish red-eye, white-eye, and skin from the intensity distribution of the eye region relating to the color of redness. In addition, since this technique requires that the eye area be specified in advance, a technique for recognizing an image of the eye area that is difficult to realize in order to be fully automated is required.

  Further, a technique for extracting a plurality of color component images peculiar to the face to detect the position of the subject's red eye, that is, extracting at least one of a low saturation region and a low illuminance region and a skin color region from the captured image, Using these extracted signals, for example, a logical product is used to extract a region including the eyes of the person, and further, red eyes are generated from the region including the eyes using the extracted red portion signal. In some cases, the red-eye position is detected, and the red at the red-eye position is corrected to another color based on the red-eye position data thus obtained, for example, a general black eye (for example, Patent Document 2). reference.). In this technique, it is not required to specify the eye area in advance, but erroneous recognition is likely to occur because the red eye position is to be recognized with the skin color and red color.

JP 2000-76427 A (paragraph numbers 0009-0011, 0058-0071, FIG. 10) JP-A-6-258732 (paragraph numbers 0006-0009, 0019-0023, FIG. 6)

  In view of the above situation, an object of the present invention is to provide a red-eye correction technique with satisfactory reliability without using a complicated image processing technique such as pattern recognition and without performing a prior work such as specifying an eye area in advance. Is to provide.

To achieve the above object, according to the present invention, red-eye correction method for correcting red eyes caused in the captured image data detects a pixel that satisfies a predetermined skin color detection condition from each pixel of the photographed image data as a skin color pixel step When the step of detecting the step of detecting a pixel that satisfies a predetermined white detection condition from each pixel of the captured image data as a white pixel, a pixel that satisfies a predetermined red detection condition from each pixel of the captured image data as a red pixel And determining the red-eye image candidate as a red-eye pixel based on a determination condition using the red color pixel as a red-eye pixel candidate and using the number of the skin color pixels and the white pixels existing in the peripheral region of the red-eye pixel candidate. And a step of correcting the red eye by changing the pixel value of the recognized red eye pixel.

  In this red-eye correction method, skin color pixels, white pixels, and red pixels are detected from the entire photographed image based on individually set detection conditions, and each red pixel is used as a red-eye pixel candidate in the peripheral region of the red-eye pixel candidate. Based on the number of skin color pixels and white pixels located, it is determined whether the periphery of the red eye pixel candidate is an eye region and the red pixel constitutes a red eye. Compared with the conventional red-eye detection method, the red-eye pixels are identified based on the skin color that characterizes the hue of the red-eye area, and the distribution of white and red. Nevertheless, it is possible to detect red-eye pixels with satisfactory reliability.

The skin color pixel and the white pixel have very similar color components (for example, R, G, B components), and it becomes difficult to distinguish them when the shooting conditions change. For example, when white is photographed in an environment of tungsten light, it becomes yellowish, so it is difficult to distinguish it from skin color. Therefore, in one preferred embodiment of the present invention, the skin color detection condition is based on the R , G, and B component values that are the pixel values of each pixel, the difference value between the R component value and the G component value, and the G component difference between the value and the B component value, B component value and a difference value between the R component values and is defined using the luminance value of the pixel, the white detection condition is a skin color of the white pixel candidates that white The skin color pixel located around the pixel candidate is defined using an average skin color saturation, and the red detection condition is defined using a red saturation. That is, the white pixel detection condition employs a value that depends on the average skin color saturation of surrounding skin color pixels. Specifically, by determining that the color displaced from the average R, G, B component values of the skin color pixels to a certain degree to the complementary color side of the skin color is white, the white color Can be distinguished.

  The red-eye correction is made by changing the pixel value of the certified red-eye pixel here, but as the simplest and most effective method of such a change, it is proposed to lower the saturation of the certified red-eye pixel Is done. By reducing the saturation, the red-eye pixel becomes grayish, and a photographic print with no sense of incongruity is obtained. Of course, it is ideal that the color of the red-eye pixel has a color balance of the actual pupil color. Therefore, when the pupil color is preset, the pixel value of the certified red-eye pixel is used as the pupil color. It is preferable to perform red-eye correction so as to approach the corresponding pixel value.

  In any case, depending on the subject, there may be a region in which white, flesh color, and red are distributed as if they were red eyes, in addition to the eye region. As a result, misrecognition of red-eye pixels occurs, and if such red-eye pixels are strongly corrected, a photo print with a sense of incongruity is returned, so if the difficulty of red-eye discrimination can be predicted in advance from shooting conditions, etc. It is meaningful to adjust the correction degree of red-eye correction according to such a degree of red-eye discrimination difficulty. For example, when shooting under special lighting, it is difficult to distinguish between colors, so the degree of red-eye correction is reduced to half or less so that even if there is a misrecognition, it does not break the overall color harmony. Good.

  The present invention also covers a program that causes a computer to execute the above-described red-eye correction method and a medium that records the program.

In the present invention, further, an image processing apparatus for implementing the red eye correction method described above are also the subject of rights, such an image processing apparatus, a skin color pixel satisfying a predetermined skin color detection condition from each pixel of the photographed image data a skin color pixel detector for detecting a pixel, the white pixel detector for detecting a pixel that satisfies a predetermined white detection condition from each pixel of the photographed image data as a white pixel, the captured image given red detection from each pixel data a red pixel detector for detecting a pixel that satisfies a condition as a red pixel, the red pixel and red pixel candidate, determination conditions using each occurrence number of the white pixels and the skin color pixels in the peripheral region of the red eye candidate pixel And a red-eye pixel determining unit that recognizes the red-eye image candidate as a red-eye pixel, and a red that performs red-eye correction by changing a pixel value of the certified red-eye pixel. And a correction unit. Naturally, such an image processing apparatus can also obtain all the effects of the red-eye correction method described above.
Other features and advantages of the present invention will become apparent from the following description of embodiments using the drawings.

Embodiments of the present invention will be described with reference to the drawings.
FIG. 1 is an external view showing a photographic printing apparatus employing a red-eye correction technique according to the present invention. This photographic printing apparatus includes a printing station 1B as a photographic printer that performs exposure processing and development processing on photographic paper P. And an operation station 1A for processing a photographed image taken from an image input medium such as a developed photographic film 2a or a digital camera memory card 2b and generating / transferring print data used in the print station 1B. ing.

  This photo printing apparatus is also called a digital minilab. As can be understood from FIG. 2, the printing station 1B pulls out the roll-shaped printing paper P stored in the two printing paper magazines 11 and prints it with the sheet cutter 12. The back print unit 13 prints print processing information such as color correction information and frame number on the back side of the photographic paper P, and the print exposure unit 14 cuts the print paper P into the size. A photographed image is exposed on the surface of the photographic paper P, and the exposed photographic paper P is sent to a processing tank unit 15 having a plurality of development processing tanks for development processing. After drying, the photographic paper P, that is, the photographic prints P, sent to the sorter 17 from the transverse feed conveyor 16 at the upper part of the apparatus is collected in a plurality of trays of the sorter 17 in a state of being sorted in order units (see FIG. 1). ).

  A photographic paper transport mechanism 18 is laid to transport the photographic paper P at a transport speed in accordance with various processes for the photographic paper P described above. The photographic paper transport mechanism 18 is composed of a plurality of nipping and transporting roller pairs including a chucker type photographic paper transport unit 18a disposed before and after the print exposure unit 14 in the photographic paper transport direction.

  The print exposure unit 14 applies R (red), G (green), and B (blue) to the printing paper P conveyed in the sub-scanning direction based on print data from the operation station 1A along the main scanning direction. A line exposure head for irradiating laser beams of the three primary colors (1) is provided. The processing tank unit 15 includes a color developing tank 15a for storing a color developing processing liquid, a bleach-fixing tank 15b for storing a bleach-fixing processing liquid, and a stabilizing tank 15c for storing a stable processing liquid.

  A film scanner 20 for obtaining photographed image data (hereinafter simply referred to as image data) from a photographed image frame of the photographic film 2a is disposed at the upper position of the desk-like console of the operation station 1A. A media reader 21 that acquires image data from various semiconductor memories, CD-Rs, and the like used as the photographed image recording medium 2b to be mounted is incorporated in a general-purpose personal computer that functions as the controller 3 of the photographic printing apparatus. The general-purpose personal computer is also connected with a monitor 23 for displaying various information, and a keyboard 24 and a mouse 25 as operation input devices used as an operation input unit used for various settings and adjustments.

  The controller 3 of this photographic printing apparatus uses a CPU as a core member and constructs a functional unit for performing various operations of the photographic printing apparatus by hardware and / or software, as shown in FIG. As described above, the functional unit particularly related to the present invention includes an image input unit 31 that takes in image data read by the scanner 20 and the media reader 21 and performs preprocessing necessary for the next processing, and various windows. And a graphic user interface (hereinafter abbreviated as GUI) for generating a control command from a user operation input (via the keyboard 24, mouse 25, etc.) through the graphic operation screen. The GUI unit 33 to be constructed and the control sent from the GUI unit 33 A print management unit 32 that performs image processing on the image data transferred from the image input unit 31 to the memory 30 in order to generate desired print data based on an operation command input from a command or a direct keyboard 24; A video control unit 35 that generates a video signal for causing the monitor 23 to display graphic data sent from the GUI unit 33, or a simulated image as a print source image, an expected finished print image, or the like during pre-judge printing work such as correction. A print data generation unit 36 that generates print data suitable for the print exposure unit 14 installed in the print station 1B based on the processed image data for which image processing has been completed, and a raw image according to customer requirements. Data or processed image data that has undergone image processing Etc. formatter 37 for formatting into a form for writing to -R and the like.

  When the photographic image recording medium is the film 2a, the image input unit 31 separately sends the scan data for the pre-scan mode and the main scan mode to the memory 30, and performs preprocessing according to each purpose. Further, when the captured image recording medium is the memory card 2b, when the captured image data includes thumbnail image data (low resolution data), the actual data of the captured image is used for the purpose of displaying a list on the monitor 23. Separately from (high resolution data), it is sent to the memory 30, but if thumbnail image data is not included, a reduced image is created from this data and sent to the memory 30 as thumbnail image data. The image input unit 31 is also connected with a self-service device called a photo print accepting device that accepts photo prints, and the print size and the number of prints are recorded from the photo print accepting device. When image attribute data and image data in which print order data and shooting conditions are recorded are received, the image data is transferred to the memory 30, and the print order data and image attribute data are transferred to the print management unit 32. In the case of a normal photo print order, print order data such as print size and number of prints, and if necessary, attribute data such as presence / absence of strobe shooting, subject information, camera type, etc. are managed by operator input through the keyboard 24. Given to part 32.

  The print management unit 32 performs an image retouching process such as color correction and filtering (such as blurring and sharpness) on the image data developed in the print order processing unit 60 and the memory 30 for managing the print size and the number of prints. A processing unit 70 is provided. In this embodiment, in particular, the print management unit 32 includes a red-eye discrimination difficulty prediction unit 80 that predicts the possibility of occurrence of red-eye and the difficulty of red-eye correction from the attribute data such as the presence / absence of strobe shooting and the input camera model. Although built, this is optional. Based on the input attribute data, the red-eye discrimination difficulty predicting unit 70 may generate red eyes (no red eye will be generated when the strobe is not used) or red-eye correction difficulty (difficulty in red-eye discrimination according to shooting conditions). However, in practice, this prediction may be left to the operator and manually input.

  The image processing unit 70 described above includes red-eye correction processing means 90 that employs the technique according to the present invention. As shown in FIG. 4, the red-eye correction processing unit 90 includes a determination condition setting unit 91 for setting and managing the skin color detection condition, the white detection condition, and the red detection condition, and the skin color from the image data based on the skin color detection condition. A skin color pixel detection unit 92a that detects pixels, a white pixel detection unit 92b that detects white pixels from image data based on the white detection condition, and a red pixel detection unit 92c that detects red pixels from image data based on the red detection condition A specific color pixel detection unit 92 including: a skin color pixel map 93a for storing the position of the detected skin color pixel; a white pixel map 93b for storing the position of the detected white pixel; and a position of the detected red pixel. A specific color pixel map 93 composed of a red pixel map 93c and a red-eye discrimination condition for discriminating a red-eye pixel from the detected red pixel. Another condition setting unit 94, a red-eye pixel determination unit 95 that recognizes the detected red-eye pixel as a red-eye pixel based on the red-eye determination condition, and a position of the recognized red-eye pixel , And a red-eye correction unit 97 that performs red-eye correction by changing the pixel value of the red-eye pixel while referring to the red-eye pixel map 96.

  The red-eye pixel determination unit 95 makes the red pixel a red-eye pixel candidate while referring to the red-pixel map 93c, and determines the number of skin-color pixels and white pixels in the peripheral region of the red-eye pixel candidate. And determining whether to recognize the red-eye image candidate as a red-eye pixel by comparing each count value with a discrimination value defined by the red-eye discrimination condition.

  The red-eye correction unit 97 that performs red-eye correction by changing the pixel value of the red-eye pixel generally lowers the saturation because the red-eye pixel becomes grayish and looks like a pupil by lowering the saturation of the red-eye pixel. Process. However, in the case of a person having a special pupil color, it may not be like an actual pupil simply by reducing the saturation. In preparation for such a case, in this embodiment, the pupil color setting unit 98 is provided, and in the red-eye correction process, the pixel value change is performed so that the pixel value of the red-eye pixel approaches the pixel value corresponding to the pupil color. It is also possible to do it.

  The procedure of red-eye correction by the red-eye correction processing unit 90 configured as described above will be described below. First, the overall flow is shown in the flowchart shown in FIG. 5, but first, skin color pixel detection (# 10), white pixel detection (# 30), red pixel from the image data developed in the memory 30. Detection (# 50) is performed. Since the detected red pixel is a candidate for red eye pixel, it is checked whether each red pixel can be recognized as a red eye pixel based on the distribution state of skin color pixels and white pixels around it, and the determination condition is satisfied This red pixel is set as a red-eye pixel (# 70). When a red-eye pixel is determined, red-eye correction processing such as saturation reduction processing and pupil color transition processing is performed on the pixel (# 90). At this time, a predetermined value or a value determined from the difficulty of red-eye discrimination is used as the degree of correction.

Next, details of each process described above will be described. The skin color pixel detection is executed in the skin color pixel detection unit 92a based on the subroutine shown in FIG. First, a skin color detection condition is fetched from the determination condition setting unit 91 (# 11). The skin color detection conditions are as follows: R, G, B luminance values of the target pixel are R, G, B, and (R + G + B) / 3 = I.
[Fs_gb_lo <GB <fs_gb_hi], [fs_rg_lo <RG <fs_rg_hi], [fs_br_lo <BR <fs_br_hi], and [I> fs_I],
It is represented by A pixel of interest that satisfies the skin color determination condition consisting of these four conditions is regarded as the skin color. Here, when the R, G, B luminance values are 8-bit color data (0 to 255), the above constants can be set as follows (hereinafter, the R, G, B luminance values are 8-bit color data). Data)
fs_gb_lo = -24, fs_gb_hi = 16, fs_rg_lo = -16, fs_rg_hi = 64,
fs_br_lo = -88, fs_br_hi = 8, fs_I = 104.
R, G, and B luminance values that are pixel values of the target pixel are taken in (# 12), whether or not the skin color determination condition is satisfied is checked (# 13), and the pixel position of the target pixel that satisfies the skin color determination condition is determined. This is stored as the pixel position of the skin color pixel (# 14). When this skin color detection is performed for all the pixels constituting the image data and the skin color check for all the pixels is completed (# 15), a skin color pixel map 93a is created based on the pixel positions of the stored skin color pixels. (# 16). The skin color pixel map 93a may be created at the step # 14 every time a skin color pixel is detected.

Next, white pixel detection is performed based on the subroutine shown in FIG. 7 in the white pixel detection unit 92b. First, the flesh color pixel map 93a created in the flesh color pixel detection subroutine is fetched (# 31), and the flesh color saturation coefficient SF is calculated from the R, G, B luminance values of the flesh color pixels to create a flesh color saturation map. (# 32). This skin color saturation coefficient SF is obtained by the following formula;
SF = SF_rg × (RG) + SF_br × (BR)
Here, by setting values such as SF_rg = 0.7 and SF_br = −1.2, the higher the saturation of the skin color pixel, the larger the value of the skin color saturation coefficient SF. The value of the coefficient SF becomes small.

Next, the white detection condition is fetched from the determination condition setting unit 91 (# 33). Further, the R, G, B luminance values of the target pixel are sequentially taken from the image data (# 34). The white detection condition is divided into two stages. First, it is checked whether or not the first stage detection condition is satisfied (# 35). This first stage detection condition is:
[Abs (GB) <ws_gb] and [abs (RG) <ws_rg] and [abs (BR) <ws_br] and [I <ws_I],
It is represented by abs () is a function expression for obtaining the absolute value of the value in (). A target pixel that satisfies the first-stage determination condition including these four conditions is regarded as a white pixel candidate. Where each of the above constants can be set as follows:
ws_gb = ws_rg = ws_br = 45, ws_I = 112.

If the target element is a candidate for a white pixel (Yes branch at # 35), the second detection condition is applied. This detection condition includes the average color of the skin color pixels in the vicinity of the target pixel for white pixel detection. Since the degree coefficient and the saturation coefficient of the white pixel candidate are included as one of the condition elements, first, the skin color saturation map created in step # 32 is referred to and 40 × 40 centered on the target pixel. An average value Ave_SF of the saturation coefficient of the skin color pixels located in the vicinity region of about the pixels is calculated (# 36). The size of this neighborhood region depends on the output gradation, and here, the output gradation is about 400 dpi. Further, the saturation coefficient SFw of the target pixel (white pixel candidate) is also calculated using the same expression as that for obtaining the skin color saturation coefficient SF (# 37). Through such pre-processing, it is checked whether the pixel of interest (white pixel candidate) satisfies the detection condition of the second stage (# 38). This second stage detection condition is:
[SFw-Ave_SF <sh_SF]
The value of the constant sh_SF is set to -12. The target pixel (white pixel candidate) that satisfies this second-stage detection condition (Yes branch at # 38) is regarded as a white pixel, and the pixel position is stored as the pixel position of the white pixel (# 39). . This white detection in two steps is performed for all the pixels constituting the image data, and when the white check for all the pixels is completed (# 40), the white pixel map 93b is obtained based on the stored pixel positions of the white pixels. Create (# 41). The white pixel map 93b may be created at the step # 39 every time a white pixel is detected.

The red pixel detection is executed in the red pixel detection unit 92c based on the subroutine shown in FIG. First, a red detection condition is fetched from the determination condition setting unit 91 (# 51). Since this red detection condition uses the red saturation coefficient SatR, first, the R, G, B luminance values of the pixel of interest are taken in (# 52), and the red, red, and red luminance values are obtained from the R, G, B luminance values as follows. A color saturation coefficient SatR is obtained (# 53).
Let min () be a function expression that takes the minimum value in () delimited by "," and max () be a function expression that takes the maximum value in () delimited by ",". ,
The coefficient mn and the coefficient mx are obtained from mn = min (G, B) and mx = max (G, B). When R> mx and R> Rsh, SatR = 100 × (R-mn) / ( R + 1) is used, and when R <= mx or R <= Rsh, the value of SatR is obtained by setting SatR = 0. Here, the constant Rsh is preferably about 50 when the R, G, and B luminance values are 8-bit color data.

  It is the red detection condition that the red saturation coefficient obtained in this way is larger than the determination constant: Sh_SatR. Here, the determination constant is set to Sh_SatR = 50, so the red detection condition is [SatR> 50. ] Is satisfied (# 54), and the pixel position of the target pixel that satisfies this red determination condition is stored as the pixel position of the red pixel (# 55). When this skin color detection is performed for all the pixels constituting the image data and the skin color check for all the pixels is completed (# 56), a red pixel map 93c is created based on the stored pixel positions of the red pixels. (# 16). The red pixel map 93c may be created at the step # 55 every time a red pixel is detected.

  The red-eye pixel is determined based on the subroutine shown in FIG. 9 in the red pixel detector 92c. First, the red-eye discrimination condition is fetched from the red-eye discrimination condition setting unit 94 (# 71). Further, since the skin color pixel information, the white color pixel information, and the red color pixel information are used for discrimination of the red eye pixel, the skin color pixel map 93a, the white pixel map 93b, and the red pixel map 93c are also captured (# 72). Since the red-eye pixel candidate pixel is a red pixel, the red-eye pixel is sequentially set as a target pixel for red-eye pixel determination (# 73). Next, a processing target peripheral area of about 40 × 40 pixels is set around the center of the target pixel (# 74). When the peripheral region is set, the number of skin color pixels and the number of white pixels existing in the peripheral region are obtained with reference to the skin color pixel map 93a and the white pixel map 93b (# 75).

  In order to make the calculation compact when calculating the number of skin color pixels and the number of white pixels in the peripheral area composed of 40 × 40 pixels, as shown schematically in FIG. It is convenient to determine the number of flesh color pixels and the number of white pixels for each block of 5 × 5 = 25 blocks. For the red pixels (indicated by A and B in the figure) located in the same block, the number of skin color pixels and the number of white pixels obtained for one red pixel are also used for the other red pixel. Therefore, it is possible to reduce the calculation load.

  The red-eye discrimination condition is that the number of skin color pixels and the number of white pixels existing in the peripheral region are equal to or greater than the respective discrimination values, and this discrimination value can be obtained experimentally. This discriminant value depends in particular on the size of the peripheral region described above, and a discriminant value that is automatically adapted is set according to the peripheral region size set in step # 74. For example, in a general human photograph, in a peripheral region composed of 40 × 40 pixels, the discriminant value for the number of flesh color pixels is 250 to 300 pixels, and the discriminant value for the number of white pixels is 15 to 30 pixels.

  It is checked whether the red-eye discrimination condition using such a discrimination value is satisfied (# 76), and the pixel position of the target pixel (red pixel) that satisfies the red-eye discrimination condition is stored as the pixel position of the red-eye pixel. (# 77). This red-eye pixel determination process is performed for all red pixels defined in the red pixel map 93c, and when red-eye discrimination for all the pixels is completed (# 78), based on the stored pixel positions of the red-eye pixels. A red-eye pixel map 96 is created (# 79). The red-eye pixel map 96 may be created at step # 77 every time a red-eye pixel is determined.

When the red-eye pixel map 96 defining the position of the red-eye pixel in the image data is created in this manner, the red-eye pixel is corrected through an appropriate red-eye correction process. In this embodiment, the red-eye correction is performed by reducing the saturation of the red-eye pixel, and the calculation formula for the saturation reduction is shown below:
d = (R + G + B) / 3,
R ′ = (1−t) × d + t × R,
G ′ = (1−t) × d + t × G,
B ′ = (1−t) × d + t × B.
That is, by changing the luminance value (R, G, B) of the red-eye pixel to (R ′, G ′, B ′), the red-eye becomes a pupil color with no sense of incongruity. Here, the coefficient t represents the degree of correction of red-eye correction. When this value is 1, no correction is made. When this value is 0, the degree of correction is maximum, that is, the color is gray. . The degree of correction: t may be set according to the degree of difficulty of red-eye correction predicted by the red-eye discrimination difficulty prediction unit 70 in some cases, and is set in advance when such difficulty cannot be predicted. A value, preferably 0.45 to 0.55, may be employed. As a result, the advantage of preventing red-eye by correct red-eye detection and the disturbance of the captured image due to erroneous detection of red-eye are balanced with each other, and a photographic print with no sense of incongruity can be obtained.

  Red-eye correction by reducing the saturation of red-eye pixels does not necessarily produce good results when the actual pupil color is light blue or green. In such a case, a process for bringing the luminance value of the red-eye pixel closer to the luminance value of the pupil color set and input by the optional pupil color setting unit 98 may be adopted as the red-eye correction process. In this case, the coefficient t is expressed as an arbitrary position on the gradation line from the luminance value of the red-eye pixel to the luminance value of the pupil color. The coefficient t is 0, no correction is performed, and the coefficient t is 1 to complete the luminance value of the pupil color. It is better to set the transition to any gradation value with the numerical value in between.

  In any case, image data that has been subjected to red-eye correction processing at an appropriate level is subjected to necessary image processing, converted to print data, and transferred to the print exposure unit 14. The print exposure unit 14 exposes the photographic paper P that finally becomes a photographic print based on the print data.

  In the above-described embodiment, the print station 1B exposes a photographed image to the photographic paper P by the print exposure unit 14 having an exposure engine, and performs a plurality of development processes on the photographic paper P after the exposure. Although the silver salt photographic printing method was adopted, of course, the printing station 1B according to the present invention is not limited to such a method. For example, an ink jet that ejects ink onto a film or paper to form an image. Various photographic printing methods such as a printing method and a thermal transfer method using a thermal transfer sheet can be employed.

  INDUSTRIAL APPLICABILITY The present invention can be widely used as a technique for incorporating into an image processing apparatus that needs to detect red-eye pixels from captured image data in which red eyes are generated.

External view of photographic printing apparatus employing red-eye correction technology according to the present invention Schematic diagram schematically showing the configuration of the print station of the photo printing device Functional block diagram explaining the functional elements built in the controller of the photo printing device Functional block diagram showing the functional configuration of red-eye correction processing means Flowchart of the entire red-eye correction process Skin color detection subroutine White detection subroutine Red detection subroutine Red eye discrimination subroutine Explanatory drawing explaining the division of the surrounding area during the red-eye discrimination process

Explanation of symbols

60: Print order processing unit 70: Image processing unit 80: Red-eye discrimination difficulty prediction unit 90: Red-eye correction processing unit 91: Determination condition setting unit 92: Specific color pixel detection unit 92a: Skin color pixel detection unit 92b: White pixel detection unit 92c: Red pixel detection unit 93: Specific color pixel map 93a: Skin color pixel map 93b: White pixel map 93c: Red pixel map 94: Discrimination condition setting unit 95: Red eye pixel determination unit 96: Red eye pixel map 97: Red eye correction unit 98 : Eye color setting section

Claims (8)

In the red-eye correction method for correcting red-eye generated in captured image data,
Detecting a pixel satisfying a predetermined skin color detection condition from each pixel of the photographed image data as a skin color pixel,
Detecting a pixel that satisfies a predetermined white detection condition from each pixel of the captured image data as a white pixel,
Detecting a pixel satisfying a predetermined red detection condition from each pixel of the captured image data as a red pixel,
Recognizing the red-eye image candidate as a red-eye pixel based on a determination condition using the red-color pixel as a red-eye pixel candidate, and a discrimination condition using each number of the skin-color pixels and the white pixels in a peripheral region of the red-eye pixel candidate;
A red-eye correction method comprising: performing red-eye correction by changing a pixel value of the certified red-eye pixel. The skin color detection condition is based on the R, G, and B component values that are the pixel values of each pixel, the difference value between the R component value and the G component value, the difference value between the G component value and the B component value, the B component value, difference value between the R component values and is defined using the luminance value of the pixel, an average skin color of the skin color pixel the white detection condition is located around the white pixel candidate with the skin color of the white pixel candidate The red-eye correction method according to claim 1, wherein the red detection condition is defined using red saturation.   The red-eye correction method according to claim 1, wherein the red-eye correction is performed by lowering a saturation of the certified red-eye pixel.   The red-eye correction method according to claim 1, wherein red-eye correction is performed by bringing a pixel value of the certified red-eye pixel closer to a pixel value corresponding to a pupil color.   The degree of change is adjusted according to a red-eye discrimination difficulty level predicted in advance when the pixel value of the certified red-eye pixel is changed for red-eye correction. Red-eye correction method. In order to correct the red eye that occurs in the captured image data,
A function of detecting a pixel that satisfies a predetermined skin color detection condition from each pixel of the photographed image data as a skin color pixel,
A function of detecting a pixel that satisfies a predetermined white detection condition from each pixel of the captured image data as a white pixel,
A function of detecting a pixel that satisfies a predetermined red detection condition from each pixel of the captured image data as a red pixel,
A function of recognizing the red-eye image candidate as a red-eye pixel based on a determination condition using each red pixel as a red-eye pixel candidate, and using the number of the skin-colored pixels and the white pixels existing in a peripheral region of the red-eye pixel candidate;
A red-eye correction program for causing a computer to execute a function of performing red-eye correction by changing a pixel value of the certified red-eye pixel. In an image processing apparatus that corrects red-eye generated in input captured image data,
A skin color pixel detection unit that detects a pixel that satisfies a predetermined skin color detection condition from each pixel of the photographed image data as a skin color pixel,
White pixel detection unit that detects a pixel that satisfies a predetermined white detection condition from each pixel of the captured image data as a white pixel,
A red pixel detector for detecting a pixel that satisfies a predetermined red detection condition from each pixel of the captured image data as a red pixel,
A red-eye pixel is determined that recognizes the red-eye image candidate as a red-eye pixel based on a determination condition using the red pixel as a red-eye pixel candidate and the number of the skin-colored pixels and the white pixels existing in the peripheral region of the red-eye pixel candidate. And
An image processing apparatus comprising: a red-eye correction unit that performs red-eye correction by changing a pixel value of the certified red-eye pixel.   A red-eye discrimination difficulty prediction unit that predicts a red-eye discrimination difficulty is provided, and the red-eye correction unit adjusts a degree of changing a pixel value of the certified red-eye pixel according to the red-eye discrimination difficulty. The image processing apparatus according to claim 7.

Images